Potential Driven Reinforcement Learning for Hard Exploration Tasks.
Enmin ZhaoShihong DengYifan ZangYongxin KangKai LiJunliang XingPublished in: IJCAI (2020)
Keyphrases
- reinforcement learning
- active exploration
- action selection
- machine learning
- multi agent environments
- search algorithm
- exploration strategy
- transfer learning
- robotic control
- database
- state space
- artificial neural networks
- learning algorithm
- data driven
- markov decision processes
- multi task
- multi agent
- exploration exploitation
- information retrieval
- model based reinforcement learning